For three decades portacaval anastomosis (PCA) represented the standard surgical therapy for cirrhotic patients with bleeding esophageal varices. Dissatisfaction with this approach could be most clearly ascertained by examining the results of three prospective controlled trials in the USA. These studies reveal that this operation provides only marginal benefit in terms of overall survival and is associated with postshunt hepatic failure and encephalopathy. The total diversion of portal venous inflow from the liver and possibly the systemic shunting of hepatotrophic pancreatic hormones have been considered to be major factors in the pathogenesis of the hepatic injury-encephalopathy syndrome. To obviate these consequences of PCA, Warren and Britton have advocated a distal splenorenal (SRS) shunt; this surgical procedure would theoretically decompress esophagogastric varices while preserving prograde flow to the liver via the portal venous system. The purpose of the present study, therefore, is to allow, by means of prospective controlled investigation, a comparative assessment of portasystemic and SRS in the management of portal hypertension. The study proposal calls for a collaborative effort of two experienced teams of investigators based at the Lemuel Shattuck-Faulkner Hospitals (Boston) and the West Haven VA Hospital (New Haven).